Efficient communication is important to every parallel algorithm. A parallel communication optimization is introduced into lattice Boltzmann method （LBM）. It relies on a simplified communication strategy which is im...Efficient communication is important to every parallel algorithm. A parallel communication optimization is introduced into lattice Boltzmann method （LBM）. It relies on a simplified communication strategy which is implemented by least square method. After testing the improved algorithm on parallel platform, the experimental results show that compared with normal parallel lattice Boltzmann algorithm, it provides better stability, higher performance while maintaining the same accuracy.展开更多
Conventional gradient-based full waveform inversion (FWI) is a local optimization, which is highly dependent on the initial model and prone to trapping in local minima. Globally optimal FWI that can overcome this limi...Conventional gradient-based full waveform inversion (FWI) is a local optimization, which is highly dependent on the initial model and prone to trapping in local minima. Globally optimal FWI that can overcome this limitation is particularly attractive, but is currently limited by the huge amount of calculation. In this paper, we propose a globally optimal FWI framework based on GPU parallel computing, which greatly improves the efficiency, and is expected to make globally optimal FWI more widely used. In this framework, we simplify and recombine the model parameters, and optimize the model iteratively. Each iteration contains hundreds of individuals, each individual is independent of the other, and each individual contains forward modeling and cost function calculation. The framework is suitable for a variety of globally optimal algorithms, and we test the framework with particle swarm optimization algorithm for example. Both the synthetic and field examples achieve good results, indicating the effectiveness of the framework. .展开更多
Communication optimization is very important for imporoving performance of parallel programs A communication optimization method called HVMP(Half Vector Message Ripelining) is presented. In comparison with the widelyu...Communication optimization is very important for imporoving performance of parallel programs A communication optimization method called HVMP(Half Vector Message Ripelining) is presented. In comparison with the widelyused vector message pipelining, HVMP can get better tradeoff between reducing and hiding communication overhead,and eliminate the communication barrier of barrier synchronization problems[1]. For parallel Systems with low bandwidth such as cluster of workstations and barrier synchronization problems with large amount of communication, HVMPmethod can get good performance.展开更多
In this paper, we present a parallel quasi-Chebyshev acceleration applied to the nonover- lapping multisplitting iterative method for the linear systems when the coefficient matrix is either an H-matrix or a symmetric...In this paper, we present a parallel quasi-Chebyshev acceleration applied to the nonover- lapping multisplitting iterative method for the linear systems when the coefficient matrix is either an H-matrix or a symmetric positive definite matrix. First, m parallel iterations are implemented in m different processors. Second, based on l1-norm or l2-norm, the m opti- mization models are parallelly treated in m different processors. The convergence theories are established for the parallel quasi-Chebyshev accelerated method. Finally, the numeri- cal examples show that the parallel quasi-Chebyshev technique can significantly accelerate the nonoverlapping multisplitting iterative method.展开更多
An optimization method to design turbine airfoils using a Genetic Algorithm (GA) design shell coupled directly with a viscous CFD (Computational Fluid Dynamics) analysis code is proposed in this paper. The blade geome...An optimization method to design turbine airfoils using a Genetic Algorithm (GA) design shell coupled directly with a viscous CFD (Computational Fluid Dynamics) analysis code is proposed in this paper. The blade geometry is parameterized and the optimization method is used to search for a blade geometry that will minimize the loss in the turbine cascade passage. The viscous flow prediction code is verified by the experimental data of cascade, which is typical for a gas turbine rotor blade section. A comparative study of the blades designed by the optimization technique and the original one is presented[展开更多
基金Project supported by the National Natural Science Foundation of China(Grant No.11002086)the Shanghai Leading Academic Discipline Project(Grant No.J50103)
文摘Efficient communication is important to every parallel algorithm. A parallel communication optimization is introduced into lattice Boltzmann method （LBM）. It relies on a simplified communication strategy which is implemented by least square method. After testing the improved algorithm on parallel platform, the experimental results show that compared with normal parallel lattice Boltzmann algorithm, it provides better stability, higher performance while maintaining the same accuracy.
文摘Conventional gradient-based full waveform inversion (FWI) is a local optimization, which is highly dependent on the initial model and prone to trapping in local minima. Globally optimal FWI that can overcome this limitation is particularly attractive, but is currently limited by the huge amount of calculation. In this paper, we propose a globally optimal FWI framework based on GPU parallel computing, which greatly improves the efficiency, and is expected to make globally optimal FWI more widely used. In this framework, we simplify and recombine the model parameters, and optimize the model iteratively. Each iteration contains hundreds of individuals, each individual is independent of the other, and each individual contains forward modeling and cost function calculation. The framework is suitable for a variety of globally optimal algorithms, and we test the framework with particle swarm optimization algorithm for example. Both the synthetic and field examples achieve good results, indicating the effectiveness of the framework. .
文摘Communication optimization is very important for imporoving performance of parallel programs A communication optimization method called HVMP(Half Vector Message Ripelining) is presented. In comparison with the widelyused vector message pipelining, HVMP can get better tradeoff between reducing and hiding communication overhead,and eliminate the communication barrier of barrier synchronization problems[1]. For parallel Systems with low bandwidth such as cluster of workstations and barrier synchronization problems with large amount of communication, HVMPmethod can get good performance.
文摘In this paper, we present a parallel quasi-Chebyshev acceleration applied to the nonover- lapping multisplitting iterative method for the linear systems when the coefficient matrix is either an H-matrix or a symmetric positive definite matrix. First, m parallel iterations are implemented in m different processors. Second, based on l1-norm or l2-norm, the m opti- mization models are parallelly treated in m different processors. The convergence theories are established for the parallel quasi-Chebyshev accelerated method. Finally, the numeri- cal examples show that the parallel quasi-Chebyshev technique can significantly accelerate the nonoverlapping multisplitting iterative method.
文摘An optimization method to design turbine airfoils using a Genetic Algorithm (GA) design shell coupled directly with a viscous CFD (Computational Fluid Dynamics) analysis code is proposed in this paper. The blade geometry is parameterized and the optimization method is used to search for a blade geometry that will minimize the loss in the turbine cascade passage. The viscous flow prediction code is verified by the experimental data of cascade, which is typical for a gas turbine rotor blade section. A comparative study of the blades designed by the optimization technique and the original one is presented[
